Duck Curve in Solar Energy

Utility companies are doing a great job but one must understand that anticipating the load demand is a critical task. Complexity comes in with the increasing production of electricity from solar energy. Curve shown below represents the load demand curve of electricity in California on 31 March 2012. The blue line in the curve represents the actual load demand in California on 31 March 2012. Least power is required over midnight. During the morning, curve ramps up when offices and schools get started, during evening time curve ramps up to the peak load and then again goes down. 

Duck Curve illustration by CAISO (California Independent System Operator) with "Net Load" versus "Time of Day" plot. Predictions for Net Load are made for different Years.

             As electricity from solar is produced only during the day time and grid gets flooded with solar energy, a dip is observed in the demand curve (red lines) during the day time with an increase in solar energy production. With further increment in solar energy production, dip observed in the curve goes on lowering down during the day time (red lines 2014-2020). The graph curve resembles a "Duck" so it was given the name "Duck Curve" by CAISO (California Independent System Operator).

              In the graph with increased solar capacity, it looks like demand has been dropped during the day time. Now the problem arises due to intense ramps in the new graph. As the sun sets, utilities have to suddenly ramp up the demand to meet the peak load which is not possible as it destroys the economics of power plant. Nuclear and Coal power plants are baseload power plants, so they need to be turned on 24X7 and can't be turned on and off every day for economic load dispatch with solar power plant. To avoid this problem curtailment of solar energy must not be considered as a solution.

Some of the solutions are:

• Integrating solar energy with energy storage technologies (batteries) could possibly eliminate the risk of overgeneration of electricity during day time.
• Switching to electric vehicles will enhance the usage of solar energy.
• Encouraging the use of innovative solutions that are cost effective such as Tesla Powerwall.
• Using Machine Learning to predict load demands much more accurately leading to lower losses.